@laughsmile: I would like to add two more aspects. Besides specific business cases and success stories there are two aspects of KNIME that seem especially relevant to me.
One is what I would call the ‘Citizen Data Scientist’ approach. Spreading KNIME within a company and see how it grows in various departments driven by a vibrant community (and maybe a core expert group). These tow examples might illustrate what I mean by that.
Five Takeaways from the First KNIME Meetup@Siemens
Empower Your Own Experts! Continental Wins the Digital Leader Award
https://www.knime.com/blog/empower-your-own-experts-continental-wins-the-digital-leader-award
The second aspect is the scalability. You could use KNIME on your laptop but you also could use it to connect to Big Data clusters and handle the analytic there. This example that runs on a local machine does also run on a large Cloudera cluster just by switching out the connectors.
Bring the latest H2O.ai models to an enterprise big data cluster (speaking of the latest machine learning …)
I recently used KNIME to teach people the usage of Hive (and partitions) on Big Data systems. It is all there in a KNIME workflow but could just like that also work on Big Data with the same nodes.
And then a further note. KNIME has an especially strong community in the Chem- and BioInformatics sector (which I know almost nothing of). That might also be relevant for your client:
https://forum.knime.com/c/special-interest-groups
I am always impressed by the workflow landscapes of @jcmozzic (which underlying science I do not understand 
)